Well, lots of problems here.
This reminds me of something I say about Sarbox, email retention, paper correspondence retention, etc. Now this article is talking about the UK, but I know this happens in the US. Considering we inherited a lot of their crappy legal system, it wouldn't surprise me if they have this too.
We have a policy of charging potential defendants with collecting, retaining, investigating, and providing potential evidence against them. Now, besides this being a braindead policy that could only come from the mind of a lawyer, it's, IMO, cleary a violation of the 5th amendment.
This, from what I know, is kind of how it works in medical studies. For example, if a local restaraunt gets a health complaint, what happens? Do they "swear" they've double checked everything and found "no violations"? And then the local health board just accepts that? Hell no, they get a visit from a health inspector. Now I'll agree that these kinds of systems are prone to abuse themselves, but any form of self-policing is just ridiculous on its face.
Yet, with a medical study, these companies certify that their drugs "pose no serious side effects". ORLY? Drugco said that? Well, then I guess Ritaxiloril must be safe right?
2nd, and this is another big obvious one that I push, abolish IP. In fact, abolish all government-granted monopolies. It has the nice side effect of abolishing all monopolies. That reduces the incentive to pull this kind of crap.
3rd, academia needs to grow a pair and throw off the journal middle men. Journals should be public domain, as well as all source data. Anything less shouldn't qualify as science, and shouldn't be treated as such.
But a world like that wouldn't leave much room for parasites, so I guess I'm just dreaming.